The identification and selection of an antibacterial chemotherapeutic agent for development depends on several properties. These include in vitro potency against bacteria, in vivo efficacy in animals and man, pharmacokinetic parameters such as good plasma levels and favorable metabolism, and reduced side effects and toxicity. The ideal agent should have the best blend of these properties.
Within the quinolone/naphthyridone class of antibacterials efforts are directed toward increasing in vitro and in vivo efficacy while lowering certain side effects such as phototoxicity and cytotoxicity and reducing general toxicity as well.
It is also known that within the chiral environment of living organisms, individual stereoisomers/enantiomers of drugs may show unique properties relative to the racemic mixtures. When this occurs, the optimal properties of the drug can only be obtained when the most favorable stereoisomer is utilized in its pure chiral form.
U.S. Pat. No. 4,665,079 shows quinolones and naphthyridones by structural formula to have 7-[3-(1-aminoalkyl)-1-pyrrolidinyl]side chains. These compounds of formula A, where R.sub.1 or R.sub.2 ##STR1## are alkyl or hydrogen were revealed to have good antibacterial in vitro potency. European Patent Publication 207,420 describes such compounds having the two asymmetric centers in the C.sub.7 side chain of the quinolone/naphthyridone and the preparation of two diastereomeric mixtures, each containing two nonseparable enantiomers of formula B and C. ##STR2##
The mixtures B and C were described to possess improved in vivo activity relative to unsubstituted compounds (Formula A, where R.sub.1 and R.sub.2 are hydrogen). All data reported were for the mixtures, and no method of separation of the mixtures was described. At the International Congress of Antimicrobial Agents and Chemotherapy (ICAAC) in Houston, Tex., 1989, there were reported certain individual enantiomers of 1-ethyl and 1-cyclopropyl-6,8-difluoro-quinolone-3-carboxylic acids. The 3-(R)-1'-(S) stereoisomers were disclosed to have the most potent activity in vitro. One stereoisomer (3R,1'S) was shown to have improved in vivo efficacy relative to an unsubstituted analog. Except for the in vitro data, no other comparisons among the pure stereoisemers were provided. The method employed to prepare and isolate the individual enantiomers involved putting a chiral auxiliary, N-tosylproline, on the amine side chain and performing a separation, removing the N-tosylproline, then replacing it with a conventional protecting group.
It has now been found that overall therapeutic value, i.e. efficacy and safety, of individual enantiomers of various 7-[3-(1-aminoalkyl)-1-pyrrolidinyl]quinolones and naphthyridones cannot be predicted until all of the enantiomers are made, separated, and tested.
It has also been found, further, that the use of a chiral auxiliary, such as (S)- or (R)-.alpha.-methylbenzylamine, as a protecting group, permits separation of diastereomeric amines, thus saving two costly steps in the overall synthesis of all four stereoisomers of 7-[3-(1-aminoalkyl)-1-pyrrolidinyl]quinolones and naphthyridines.